The Internet is a dispersed collection of computers communicating with each other using a protocol called TCP/IP (Transmission Control Protocol/Internet Protocol). These computers fall into two categories: "host" computers, also referred to as servers, and "client" computers. Host computers form the backbone of the system and each is connected by a dedicated link to at least its nearest neighbouring host. Host computers are typically owned and operated by large governmental, academic and commercial organisations to provide services to their members and clients. Although the Internet grew out of academic and government networks an increasing proportion of the servers in it are operated by "Service Providers" who provide access to the general public for a fee.
Client computers form the interface between users and the hosts or servers. They are connected to the servers either directly, by a local area network (LAN) or by dial-up telephone line and modems. The latter is usual for members of the public who purchase access from service providers.
Host computers are identified by a four part IP address in the form AAA.BBB.CMDDD, where each component is a number less than 256, which are related to more user-comprehensible names by the Domain Name Server (DNS) system. For example, one host at Sony Europe is called "opus.ipek.de", which corresponds to the IP address 194.77.200.3. "opus" is the host name of that machine and "ipek.de" is the domain name of the IP address range 194.77.200.1 to 194.77.200.62. Each resource on the Internet, including Web pages (see below), has a unique address known as a Uniform Resource Locator (URL) which includes the hot name and domain name of the server hosting it. Specific users are identified as "userid"@ "hostname.domainname", for example "FredBloggs@www.sony.com".
When a host computer has a message or address for another host it sends it to a host to which it has a direct connection that is in the general direction of the ultimate destination of the message or request. That computer then forwards the message to another host again in the general direction of the ultimate destination and so on until the message is delivered.
The Internet provides a number of services to its users of which the most popular is the World Wide Web (WWW). The World Wide Web is essentially a collection of files, often referred to as Web pages, in a variety of formats stored on host computers, often called Web servers, in the Internet. A collection of Web pages published by an organisation is often termed a Web site and its first or highest level page the homepage. Most Web pages are textual documents described in Hypertext Mark-up Language (HTML). Such pages may include images, still or moving, and audio data. The key feature of HTML is the ability to define Hypertext Links within the document which provide access points to other parts of the same document, other Web pages or other Internet facilities.
To access the World Wide Web, the user employs software on his client computer known as a Web browser, commercially available examples are Netscape Navigator.TM. and Microsoft Internet Explorer.TM.. The user enters the address of a Web page, or selects one from a list of previously stored addresses often referred to as Bookmarks. This address includes the address of the Web server holding that Web page. The Web browser then automatically contacts the user's service provider, dialling up a link over the telephone network if necessary, and issues a request for that Web page. This is forwarded to the relevant host computer which replies by sending back the relevant page, if it exists. The communication of the Web page is conducted using the Hypertext Transfer Protocol (HTTP) as a result of which World Wide Web Page addresses begin "HTTP://". Access to most Web pages is unrestricted however it is possible for this to be controlled by use of passwords.
As the Web page is received by the client computer the user's Web Browser software constructs and displays the page according to the instructions specified in HTML. This may allow a degree of control to the user's Web browser software or may precisely specify the locations of the components in the displayed image. Hypertext links comprise two components, a part to be displayed to the user, either text or an image such as an icon, and the associated address, which may be within the same Web page or may be a completely different page. Different addresses may also be associated with different parts of an image. The part of the link displayed to the user is generally distinguished from other parts of the page, for example in the case of text by being underlined or in a different colour. Generally the client computer will use a mouse (or trackball) and on-screen pointer for input of commands, in this case the pointer is often arranged to change shape or colour when located over a hypertext link. When the user selects a hypertext link, usually by positioning the pointer over it and clicking a mouse button, the Web Browser software automatically accesses the corresponding Web page.
The World Wide Web is extremely popular and continually growing but users often complain that it is slow to access the Web pages in which they are interested. Whilst a user may have a relatively fast link to his service provider, e.g. a 28.8 kbps modem, the connection from the service provider to the server containing the Web page being loaded may involve several different link hops shared with other traffic so that the effective speed of this is considerably less, especially at peak times.
Most users accumulate collections of Web pages that they visit often. Accessing these pages is assisted by the provision of bookmarks, which are links to the corresponding pages pre-stored in the user's Web browser. However many pages are updated at unpredictable intervals at the whim of the author. The user cannot tell whether or not a Web page has changed without visiting it, which can be a time consuming process.
In part because of the lack of a secure method of arranging payments over the Internet, most organisations providing Web pages and discussion forums do not charge users for access. Instead, some attempt to recuperate their costs by displaying advertising. Some service providers also attempt to keep their charges down or make additional profit by carrying advertising. Such advertising delays access to the Web pages or other services and can be particularly irritating to the user if he or she is paying for access time by the hour and also paying for the telephone connection.
To advertise their services, many Web sites include "What's New" listings in their homepage. Some service providers also provide such listings each time the user accesses the service. These can again be irritating to the user as they take time to transmit for which the user is paying.
As far as the providers of Web sites are concerned, to maintain a Web site generally requires a knowledge of HTML and often considerable input from the site administrator. Software to assist in this is available but does not completely dispense with the need for skilled input. This hinders the regular updating of Web pages by unskilled users.
Whilst the Internet is a global open network, its advantages and ease of use have been recognised as useful for smaller, closed networks. Networks using the Internet's protocols and software but confined to a single organisation are termed intranets. They suffer from the same problems as the Internet to varying degrees.